Hi,
When a job is abruptly terminated in the middle of a calculation,
a foobar_STATUS is left behind.
Can I use this STATUS file to easily restart the calculation
more-or-less from the place where it was stopped?
If yes, then how can I do that?
Is this documented somewhere?
Thank you!
Rob.
Can I continue interrupted calculation with STATUS file?
Moderator: bguster
-
spamrefuse
- Posts: 35
- Joined: Wed Jan 20, 2010 3:08 am
Re: Can I continue interrupted calculation with STATUS file?
The STATUS file is not used to restart a calculation.
The following files may be helpful:
- _WFK, contains the wavefunctions, big file, can be used to restart exactly the calculation at the point it stopped previously, see prtwf and irdwfk and getwfk variables;
- _DEN, contains the density, smaller file, can be used to establish the potential for the initial step of SCF loop, may be faster to help the self consistency like that, see variables prtden and friends;
If you're doing a geometry relaxation, when restarting, don't forget to cut and past the last coordinates from the log (or output) to the new input restart file for your atomic coordinates to match with your WFK file.
Damien.
The following files may be helpful:
- _WFK, contains the wavefunctions, big file, can be used to restart exactly the calculation at the point it stopped previously, see prtwf and irdwfk and getwfk variables;
- _DEN, contains the density, smaller file, can be used to establish the potential for the initial step of SCF loop, may be faster to help the self consistency like that, see variables prtden and friends;
If you're doing a geometry relaxation, when restarting, don't forget to cut and past the last coordinates from the log (or output) to the new input restart file for your atomic coordinates to match with your WFK file.
Damien.